dbt faker

Generate fake/test/demo/sample data directly from dbt. dbt_faker is a python model generator for generating data within a dbt project using the Python Faker project.

Install

Include in packages.yml:

packages:
  - git: "https://github.com/dbt-labs/dbt_faker.git"
    revision: main 

Requirements

• dbt version >= 1.3

How to use it

1. Add a source override macro in your project

Create the file macro/dbt_faker_source_override.sql that looks like this:

{% macro source(source_name, table_name) %}
{{ return(dbt_faker.dbt_faker_source(source_name, table_name)) }}
{% endmacro %}

Activate the faker_enabled variable in your project.yml

vars: 
  faker_enabled: true

2. Declare your sources.yml

including columns and faker_providers, and add the meta config faker_enabled:true.

version: 2

sources:
  - name: tpch
    meta:
      faker_enabled: true
  - name: fake_tpch
    tables:
      - name: orders
        meta:
          faker_enabled: true
          faker_rows: 250
        columns:
          - name: o_orderkey
            meta:
             faker_provider: pyint

          - name: o_order_date
            meta:
              faker_provider: date

3. Generate your python model

Execute the command dbt run-operation generate_faker_model

4. Copy the output of your terminal and create a python model

Create a file (e.g. dbt_faker.py) with the code generated from step #2

5. Execute your newly created python model

For example dbt run -m dbt_faker.py. This will create a table called fake__source_table for each source you have defined as fake-able

6. Use your fake data!

Run the models depending on the fake sources and be amazed

Providers

dbt_faker relies on Faker's robust data providers. In order to use them, simply include the name of the provider in the faker_provider meta tag. A full list of providers is here. Some examples you can use:

• faker_provider.address (48764 Howard Forge Apt. 421 Vanessaside, PA 19763)
• faker_provider.name ( Diego Maradona)
• faker_provider.pyint (1234)

If a fake_provider has not been defined for a column, dbt faker will generate a string by default.

FAQ

generate_faker_model is skipping my sources

You should check that your sources have: 
    - Columns defined in the sources.yml
    - the meta field faker_enabled: true either at the source name level or source table name level
    - the meta field faker_enabled:false not defined at the source table level 

dbt run -m dbt_faker.py gives me a warning that the selector haven't found the model

You may not be running dbt 1.3, needed to be able to execute dbt python models